Chronicle of the seminar “Attributive philology and digital analysis of literary texts” (part 1)
On December 10 and 11, 2020, the two-day seminar “Attributive philology and digital analysis of literary texts” was held online by Professors Pasquale Stoppelli (formerly “Sapienza” University of Rome) and Paolo Mastandrea (Ca’ Foscari University of Venice). The meeting, which was attended by a total of 270 people (174 on the first day and 96 on the second), is part of the cycle “New horizons. How to renew the humanities” (HFC-HOR 2020), organized by Humanities for Change, promoted by the Venice Centre for Digital and Public Humanities (VeDPH) and financed with funds for student activities of Ca’ Foscari University of Venice.
The first day took place on Thursday, December 10 from 2 to 4 p.m. and started with the institutional greetings of Professor Giovanni Vian (Director of the Department of Humanities of Ca’ Foscari University of Venice), followed by a talk by Professor Franz Fischer (Director of the Venice Centre for Digital and Public Humanities); both of them, after thanking Humanities for Change, stressed the importance of the international network within the so-called “third mission”.
The introduction by Marco Sartor (University of Parma) illustrated the aims of the meeting, intending to stimulate a reflection on the changes that have affected philological practice in the digital age and on the added value of the interrogation of textual databases for the purposes of exegesis and philology of literary texts. Questions already raised in the middle of the last century with Father Roberto Busa (1949), considered by most as the father of digital humanities, and propagated also thanks to Gianfranco Contini’s exhortation to concordances (1951). Since then, many steps forward have been taken: various electronic corpora were born (and in some cases even died, due to the obsolescence of the computer technologies with which they were produced) , such as Letteratura Italiana Zanichelli (LIZ), Musisque Deoque (MQDQ), Tesoro della Lingua Italiana delle Origini (TLIO), Archivio della Latinità Italiana del Medioevo (ALIM), and so on. The scrutiny of electronic indexes, Sartor underlines, can play a decisive role in the attributive methodology of literary texts and Pasquale Stoppelli discusses this aspect in the first chapter of the recent volume L’equivoco del nome. Rime incerte fra Dante Alighieri e Dante da Maiano (Roma, Salerno Editrice, 2020). If the proposal originates from an intuition, in order to be verified, it needs evidence to corroborate or disprove it. It is precisely at this delicate stage that electronic repertories play a substantial role.
Pasquale Stoppelli and the attribution of the Fiore
Sartor’s introduction was followed by Pasquale Stoppelli’s talk, focused on the problem of the attribution of the Fiore, a crown of 232 sonnets whose date of composition, although uncertain, seems to be between the thirteenth and fourteenth centuries. The most direct source of the text is the Roman de la Rose, a poem in oïl language dating from the second half of the 13th century. The story describes a love conquest in which various characters interact, some siding with the Lover, others defending the flower itself, that is, the Beloved. The only codex bearing the text, manuscript H 438 of the Inter-University Library of Montpellier (datable to the 1440s), is anonymous and anepigraphic. The author declares himself twice in it with the name of Ser Durante, an element that led the first editor (Ferdinand Castets) to formulate the hypothesis of Dante’s authorship of the Fiore. The proposal sparked a lively discussion that saw among the interlocutors also Gianfranco Contini, who in 1984 republished the work at the Società Dantesca arguing vigorously the same thesis of Castets and citing linguistic and stylistic comparisons with the Supreme Poet. With the volume Dante e la paternità del Fiore (Roma, Salerno Editrice, 2011), Stoppelli fits into the debate by proposing a new research methodology that takes shape through a sort of “radiography” of the poem, sonnet by sonnet, through the use of digital databases (in particular LIZ and TLIO).
This preamble has given way to the presentation of practical examples, with the aim of highlighting the abundance of concordances that seem to connect the poem with the compositions of different rhymesters, including Monte Andrea, Guido Cavalcanti, Lapo Gianni, Brunetto Latini, Dante da Maiano and, obviously, Dante Alighieri himself. In the light of such data, Stoppelli considers it unlikely that the poets mentioned above were all readers of the Fiore, which would not have had a considerable circulation in Tuscany (it is not by chance that the tradition is uni-testimonial). Moreover, the centonary practice that distinguishes the crown of sonnets is a mode that is not very suitable to Dante Alighieri’s poetry. Therefore, Stoppelli reflects on the possibility that his identification as the author of the Fiore derives from the ease with which a certain “cultural” memory tends to remember Dante’s quotations and to neglect, conversely, those of his preceding and subsequent authors, which could only be detected through the use of digital textual databases.
The conviction of Dante’s authorship has necessary repercussions on the ecdotic level, since, starting from this assumption, the editor will not be able to accept stylistic features incompatible with the poetics of the Supreme Poet and will proceed with the introduction of the necessary amendments. The example given by Stoppelli concerns in particular vv. 9-12 of sonnet 18:
Bellacoglienza disse: “I’ vo’ che vegna, e basci il fior che tanto gli è [’n] piacere, ma' ched e’ saggiamente si contegna; ché siate certa che no·m’è spiacere”.
Contini’s amendment concerns an original piacente amended in piacere (v. 10), so that the rhyme scheme is respected. However, Stoppelli points out that if the author was a Guittonian, the lesson could be accepted because it would be related with saggiamente in the following verse to form an internal rhyme. The second problem would concern the verb siate (v. 12) which, considered trisyllable by Dante onwards, would give rise to a hypermeter verse. The use of databases has allowed to detect a use of the verb siate bisyllable in the lauda 34 of Iacopone da Todi (and therefore in pre-stilnova poetry). Here, therefore, those loci considered errors by Contini would be, on the contrary, the proof of a poetry prior to Dante’s production.
Once Professor Stoppelli’s presentation had ended, Sartor began the discussion by asking a question concerning the role of databases in relation to some issues related to the “tenzone del Duol d’amore” between Alighieri and Dante da Maiano, published in 1527 in the “Giuntina di Rime antiche”, and which the scholar discusses in the already mentioned L’equivoco del nome. Rime incerte fra Dante Alighieri e Dante da Maiano. In particular, the problem would concern the fictitious character of the tenzone sustained by Stoppelli, who considers it a staging by the minor Dante who, in order to accredit himself, would have faked a dialogue with his namesake.
Stoppelli’s answer underlines the centrality of the role of digital research, bringing it back to the theme of intuition, understood as original thought, innovative illumination that can contradict – in whole or in part – an entire tradition of studies, in the presence of which it is necessary to place oneself with deferential respect but also free from the certainty that it has absolute value. If the role of the philologist, says Stoppelli, is to increase (and not rehash) the information on a given text, the identification of new objective data can be done in a practical and immediate way through digital databases. Their use has therefore provided the philologist with data that, on the one hand, would deny Dante’s authorship of part of the Duol d’Amore and, on the other hand, would lead to the hypothesis that Dante da Maiano is the author of the Fiore, by bringing out striking analogies between the tenzone and the poem.
The dialogue continued with a question on how databases can be used in a way that is not superficial but fruitful, underlining the reluctance of the academic world to teach courses on how to use them in practice. Professor Stoppelli’s answer indicated a hypothesis as a logical-operational necessity, a useful question to be submitted to the machine, without, however, constraining the research in any way. The interactive investigation has in fact a heuristic value, being able to suggest answers to which the researcher had not thought of: hence the scientific progress. At this point, the speaker underlined today’s difficulty in creating or enriching databases as a result of increasingly rarefied funding, expressing admiration for the Opera del Vocabolario Italiano (OVI), which continues its project with great effort thanks to the help of the CNR.
This was followed by an intervention by Matteo Veronesi who raised the problem of the difficulty of databases in detecting non-lexical concordances,, to which Sartor replied by citing both the example of Musisque Deoque, capable of recognizing phonic and metric-verbal occurrences and therefore the emergence of examples of unconscious memory, and the case of Memorata Poetis, which is currently being updated with regard to semantic research. Pasquale Stoppelli, on the other hand, argued the need for the emergence of second-generation databases equipped with sophisticated software capable of recognizing conceptual associations and extracting a large amount of information from a lightly encoded text. For now, they can only be retrieved by searching for strings of characters representative of a given concept. Veronesi, therefore, returning to the main theme of the Fiore, draws attention to an argument put forward by supporters of Dante’s authorship, namely the fact that the candid rose of Dante’s Paradise may appear, on a general hermeneutic level, to be a sublimation and palinody of the very carnal and profane rose of the early poem. With the second question, instead, Veronesi asks if the versification expertise and the centonary technique cannot be explained by the young age of the poet and the experimental nature of the exercise, a historical fact that the literal analysis ignores. To these thorny questions Stoppelli contrasts various chronological data that would not seem to lead to a young Dante, who at the age of twenty had already written the Vita Nova, but rather to an author lacking the awareness of the boundaries between the two languages – Vulgar Tuscan and Old French – and to a dating between the ’20s and ’40s of the fourteenth century.
Paolo Mastandrea: databases as telescopes
This is where Paolo Mastandrea’s contribution comes in, opening with an appeal to the academic world of the humanities to take on the task of carrying, in the delicate era of the transition from paper to digital, a quantity of knowledge costing generations. The potential of databases is compared by analogy to that of the telescope, which objectively shows elements invisible to the naked eye; in this regard, an example is given of the first three octaves of Gerusalemme Liberata. Although the proximity of the expression
col senno e con la mano (v. 3)
to the Ovidian expression
consilioque manuque (Met. XIII, 205)
col senno e con la spada (Inf. XVI, 39)
and Ariosto’s expression
col senno e con la lancia (O.F. III, 55, 2)
had been noted by previous commentators, the use of digitized libraries has made it possible to unearth what was probably the model for the elegiac, namely this expression from Ilias Latina:
animoque manuque (v. 712)
The discussion is brought back to the theme of the attribution of the Fiore by the discussant Irene Mamprin, who reflects on the etymology of the name Durante, that is “the one who perseveres”: couldn’t it be considered an allegory like Malabocca or Ragione rather than a real person? Secondly, he mentions some cases of doubtful attribution in Latin, in particular the Donatists texts and the Historia Augusta, and asks Professor Mastandrea if it is possible to solve these problems through the use of new digital means. Stoppelli’s answer brings out an element that would seem to contradict this hypothesis, namely the fact that the name is placed at the same height as that of Jean de Meun in the Roman de la Rose. Mastandrea, on the other hand, argues that, although at the time of his research on the Donatists texts the capabilities of computers were not as sophisticated as they are today, he was able to contradict the belief that the Passio Massimiani and the Passio Mercuri were by the same author because the former presented only one case of use of a specific emphatic form (an uncontrolled use of superlative adjectives), while the latter twenty-three. With regard to the Historia Augusta, however, the speaker emphasizes a resumption of interest through computational studies.
At this point Alessia Luvisotto shifts the chronological axis to contemporary literature with a question about one of the most interesting and discussed cases of attributionism of the 20th century, namely Eugenio Montale’s Diario Postumo, which some scholars attribute to Annalisa Cima: is the methodology used to demonstrate the non-paternity of Dante’s Fiore usable also for texts of different historical periods (and therefore conceived in historical and cultural contexts that provide for a different circulation of literary texts)? How can we face, for contemporary literature, the problem of copyrights that prevent the creation of virtual libraries that can be freely consulted? Pasquale Stoppelli replies to these questions with confident sentences: the methodology can be the same, paying attention to calibrate the research according to the characteristics of the single poet, of the genre, of the epoch in which the work was born. Copyright, a legitimate recognition of the work of a given author, does not prevent research, as it is always possible to create private databases. The problem with contemporary poetry is the substantial lack of this kind of research tools, which are present and complete for ancient and modern texts.
Nicola Cittadini’s subsequent intervention questions the risk that, in the interpretation of objective data, the philologist, jealous of his own iudicium, may shift the place of evaluation to the findings of the databases, as Bedier stated for the bipartite trees.
Judgment is essential in research and progress, especially in literary subjects, where everything can be ambiguous.Pasquale Stoppelli
With the quotation of Federico Zeri’s book, Due dipinti, la filologia e un nome (Milano, TEA, 1995), Michele Lodone introduces the theme of attributionism in figurative works of art, which is welcomed by the speaker with the reference to Giovanni Morelli and his method to attribute dubious pictorial works: one must examine not the whole or the critic’s feeling, but the apparently insignificant details. The philologist will have to do the same: there is no need for appreciable or enjoyable critical formulas, but for philological evidence, objective data.
At the end, after a question from Sartor on the usefulness of databases in detecting forgeries, followed by Stoppelli’s clarification on the necessity to always interrogate the machine properly, the first session of the seminar ended with greetings and thanks from Irene Mamprin.
To learn more
- Contini, Gianfranco (1985). Il Fiore e il Detto d’Amore attribuibili a Dante Alighieri. Milano: Mondadori.
- Stoppelli, Pasquale (2011). Dante e la paternità del Fiore. Roma: Salerno Editrice.
- Stoppelli, Pasquale (2020). L’equivoco del nome. Rime incerte fra Dante Alighieri e Dante da Maiano. Roma: Salerno Editrice.
- Zeri, Federico (1995). Due dipinti, la filologia e un nome. Milano: TEA.
The English translation of this article has been revised by Chiara Marabese.